Negative Thermal Expansion Materials

 J. D. Jorgensen

This work was begun as a collaboration with A. W. Sleight (Oregon State Univ.) and eventually grew to produce a number of publications.  A renewed interest in negative thermal expansion (NTE) occurred when Sleight showed that ZrW2O8 exhibits a very large isotropic NTE over the temperature range 4-1050 K. Previously studied materials exhibit bulk NTE over a limited temperature range or NTE along one crystallographic axis. Our work began with the study of ZrW2O8 under pressure to determine what kinds of structural changes could give rise to volume reduction. This work was extended to studies as a function of temperature and studies of chemically modified compounds.

Contribution of Structural Fluctuations to Negative Thermal Expansion.  Using in situ neutron diffraction studies at high pressure we learned what changes in the structure, i.e., rearrangements of the atoms, can reduce the cell volume.  We then showed that, at ambient pressure, these same structural changes occur as thermally-excited defects, or fluctuations dispersed randomly in the structure, and contribute to the NTE.  The fluctuations are made possible by the thermally-excited migration of oxygen atoms over a path of about 4 �.  This novel mechanism for NTE, which was predicted from statistical mechanics and thermodynamics in the late 1950's (by L. D. Landau and E. M. Lifshitz) has previously been used to explain the NTE of water in a very small temperature range just above its melting point; but, it has not previously been reported in a solid.

Optimizing the NTE Materials for use in Composites.  In our study of ZrW2O8 under pressure, we learned that the materials undergoes a first-order structural phase transition, with a 5% reduction in cell volume, at the low pressure of 0.2 GPa. When ZrW2O8 is used as a component to control the thermal expansion of a composite, the grain interaction stresses can be large enough to create local pressures that drive this transition.  Thus, such a composite exhibits strange irreversible behavior as the ZrW2O8 particles go back and forth through this transition with considerable hysteresis.  To investigate solutions to this problem, we studied chemically-modified compounds with the same structure.  HfW2O8 was found to undergo the same phase transition, but at a significantly higher pressure -- 0.6 GPa.  ZrMo2O8 does not undergo a phase transition at any pressure up to 0.6 GPa.  Either material is, thus, suitable for use in composites within their temperature ranges of stability.